Benchmark Almanac

CBT-Bench Level II

Claude Opus 5.5 0.63 leads since September 2026
Not in model cards

Models identify cognitive distortions and classify patients’ core beliefs at two levels of detail. The evaluator compares the selected labels with the reference labels.

Slingshot AI ranks ten models by mean weighted F1 across three classification tasks. Its comparison covers Level II of CBT-Bench.

Made by
Mian Zhang and colleagues
Released
October 2024
Kind
Medicine
Level
Expert
Scored
Checked automatically
Answers
Multiple choice
Tasks
All public
Measures
Capability
Language
English
In and out
Text → Text
Interaction
Single turn
Refresh
One fixed set
Run by
Anyone
Strongest model tested
GPT-6 Astra #4 on the capability index
Last model added
September 2026
Scores refreshed
November 20th, 2026
10
Models scored on Slingshot AI’s leaderboard

Claude Opus 5.5 leads with 0.63 among models released by September 2026

Best score per model from Slingshot AI’s leaderboard, read on November 20th, 2026. Where a model ran with several settings or harnesses, we keep its best.

The newest model with a score here came out in September 2026, and nobody has tested a newer one. Among the 20 highest-ranked models on the capability index, these have no score yet:

  1. 1 Claude Opus 5.5 temperature=0, max_tokens=2048 0.63
  2. 2 Gemini 2.5 Pro Jun 2025, temperature=0, max_tokens=2048 0.63
  3. 3 Gemini 4 Argon temperature=0, max_tokens=2048 0.62
  4. 4 GPT-5 temperature=0, max_tokens=2048 0.62
  5. 5 Qwen3.5-122B-A10B temperature=0, max_tokens=2048 0.61
  6. 6 GPT-6 Astra temperature=0, max_tokens=2048 0.61
  7. 7 GPT-6 Luna temperature=0, max_tokens=2048 0.60
  8. 8 GPT-4o Aug 2024, temperature=0, max_tokens=2048 0.60
  9. 9 Claude Sonnet 4.5 temperature=0, max_tokens=2048 0.60
  10. 10 Qwen3.5-27B temperature=0, max_tokens=2048 0.59

Source: Slingshot AI’s leaderboard

#ModelScoreSettingDateLinkSource
1 Claude Opus 5.5 0.63 temperature=0, max_tokens=2048 Sep 2026 openrouter.ai Official leaderboard
2 Gemini 2.5 Pro Jun 2025 0.63 temperature=0, max_tokens=2048 Jun 2025 openrouter.ai Official leaderboard
3 Gemini 4 Argon 0.62 temperature=0, max_tokens=2048 Sep 2026 cloud.google.com Official leaderboard
4 GPT-5 0.62 temperature=0, max_tokens=2048 Aug 2025 developers.openai.com Official leaderboard
5 Qwen3.5-122B-A10B 0.61 temperature=0, max_tokens=2048 Feb 2026 huggingface.co Official leaderboard
6 GPT-6 Astra 0.61 temperature=0, max_tokens=2048 Sep 2026 developers.openai.com Official leaderboard
7 GPT-6 Luna 0.60 temperature=0, max_tokens=2048 Sep 2026 developers.openai.com Official leaderboard
8 GPT-4o Aug 2024 0.60 temperature=0, max_tokens=2048 Aug 2024 developers.openai.com Official leaderboard
9 Claude Sonnet 4.5 0.60 temperature=0, max_tokens=2048 Sep 2025 openrouter.ai Official leaderboard
10 Qwen3.5-27B 0.59 temperature=0, max_tokens=2048 Feb 2026 openrouter.ai Official leaderboard

Rankings by part

We read 16 parts from Slingshot AI’s leaderboard in 3 groups: task, cognitive distortion (f1 per label) and primary core belief (f1 per label). Each ranks models on one part of the test. Gemini 2.5 Pro leads 7 of them, and 6 other models lead the rest. Arrows show how far a model moved from its overall rank among the models we track.

Cognitive distortion

10 models we track have a score here.

  1. 1 Gemini 2.5 Pro Jun 2025 ↑1 0.50
  2. 2 Qwen3.5-122B-A10B ↑3 0.48
  3. 3 GPT-5 ↑1 0.48
  4. 4 Claude Sonnet 4.5 ↑5 0.48
  5. 5 Gemini 4 Argon ↓2 0.47
  6. 6 Claude Opus 5.5 ↓5 0.47
  7. 7 GPT-6 Astra ↓1 0.47
  8. 8 GPT-6 Luna ↓1 0.46
  9. 9 Qwen3.5-27B ↑1 0.45
  10. 10 GPT-4o Aug 2024 ↓2 0.44

3 models have held number one on Slingshot AI’s leaderboard since its release in October 2024

Lab claims count from the day the card came out, independent runs from the model’s release.

OpenAI Google DeepMind Anthropic
Slingshot AI leaderboard Not tested since Sep 2026

Source: Slingshot AI’s leaderboard

Scores over time

All the scores we have, one dot each, so a model tested at several settings appears more than once.

Plot with other benchmarks
On Slingshot AI’s leaderboard
0.60 0.62 0.64 2025 2026 0.60 0.62 0.64 2025 2026

Source: Slingshot AI’s leaderboard

No lab model card reports it yet

We read the headline results of 540 launch posts from 57 labs, and none of them list this benchmark.

More in Law, finance and medicine

All 637 in Law, finance and medicine
  • MedXpertQA-MM In use Models answer expert-level medical exam questions that combine images with clinical records and test results. Scoring is accuracy. Qwen 3.8 Max 80.4%
  • HealthBench Professional In use 525 physician-authored tasks from real clinician chats (care consult, documentation, research) with physician-adjudicated rubrics. Claude Sonnet 5.5 69.2%
  • HealthBench In use 5,000 realistic health conversations graded against physician-written rubrics. A length-adjusted variant penalizes verbosity. GPT-5 67.2%
  • FinSearchComp In use Agents search the web to fetch live financial data, look up history, and run multi-step investigations, across 635 expert-written questions. Scored by answer accuracy. Ling-3.1-flash 81.7%
  • KorMedMCQA In use Models answer 7,469 multiple-choice questions from Korean licensing exams for doctors, nurses, pharmacists and dentists, held from 2012 to 2024. o1-preview 92.7%
  • Harvey LAB-AA v1.1 In use Artificial Analysis's run of Harvey's Legal Agent Benchmark on Harvey's newer private set of 120 tasks across 24 practice areas. A deliverable with a material hallucination scores zero, and the headline is the share of judges who find every rubric criterion passed. GPT-6.2 Astra 15%

Data audit

All 10 checks pass. Every benchmark

Catalog record 2 pass
  • Pass Task access All tasks are public.
  • Pass Confidence Checked against a primary source.
Scores 4 pass 1 note
  • Pass Independent runs Scores from the official leaderboard (10 entries).
  • Pass Models scored 10 models we track.
  • Pass Newest score The newest is from September 2026, for Gemini 4 Argon.
  • Pass Frontier models 3 of the top 10 have a score, led by GPT-6 Astra (#4).
  • Note Headroom Scored as a plain number, with no fixed ceiling.
Leaderboard import 3 pass
  • Pass Importer We read the board on Slingshot AI: 10 rows for models we track at the last read.
  • Pass Last read November 20th, 2026.
  • Pass Read worked The last read worked.
Consistency 1 pass 1 note
  • Note Lab claims No model has both a lab number and an independent run.
  • Pass Known issues No published audit found problems.

How we check